Loading ...
Sorry, an error occurred while loading the content.

Position Papers ...

Expand Messages
  • Theo D'Hondt
    There s nothing new under the sun ... Until somebody figures out how to tap the power of the multiverse and harness a gaggle of qubits into a semblance of a
    Message 1 of 7 , Apr 10 3:16 AM
    • 0 Attachment
      "There's nothing new under the sun ..."

      Until somebody figures out how to tap the power of the multiverse and
      harness a gaggle of qubits into a semblance of a computer I am going
      to assume that the universal Turing machine is what we're talking
      about here. Therefore I choose to interpret "resetting computation"
      as identifying a significantly better permutation of the
      computational wisdom (for want of a better description) that we've
      been accumulating for the past 70 years. There's no shame in doing
      this: after all, people didn't stop composing music in 1750.

      Picking up the UML+XML+JAVA theme (and let's add the Java VM for good
      measure) I would suggest that freezing a list of distinct language
      formalisms to express views on the same content points to lousy
      science - what happened to the 0-1-infinity rule? Unfortunately,
      lousy science seems what we're heading for.

      However, in the various e-mails I see a lot of references to
      fundamental philosophical issues with computation, programming,
      building systems, &tc. I would happily join this discussion (in
      particular the one viewing software construction as a Heisenbergian
      interaction ... ) but I fear we have a much more immediate and down
      to earth task.

      I seems to me that we are uniting a group of people that are
      extremely knowledgeable about this accumulated computational wisdom.
      A consolidated effort to rearrange at least part of this knowledge in
      recommendations for a "reset in computation" might not seem very
      challenging to some of you; on the other hand it might prove to be
      extremely useful. Call me naive, but it seems to me that we might
      make more of a splash with a pragmatic approach than with a
      philosophical one.

      So I beg to differ with some of you - I think our discussion should
      include programming technology such as MOPs, reflection &tc. I might
      be old fashioned but I still think an expressive programming
      environment is a software developer's best friend. Expressiveness is
      the single most important attribute of any language and things
      started going wrong when this principle was discarded (forgotten?) by
      the software industry.

      Many of us are educators and our educational concerns have an
      important place in this discussion. I tend to share Dave Thomas'
      preference for a focus on competence, maturity and knowledgeability
      in development groups; this stands foursquare in opposition to the
      process approach to software (which I fear is the mythical man-month
      by another name). The latter is certainly inspired by the typical
      profile of a computer professional today: his or her historical
      perspective and historical consciousness is generally non existent.
      We can only blame the educational system (and therefore ourselves)
      for this state of affairs where education becomes training and is
      driven by immediate concerns of the industry.

      So I would like to propose a two-stage approach to the "reset in
      computation". First of all I would to see us all pool our knowledge
      of the field of computation, and try to marshall the results of the
      past into a coherent plan for the future. I am deliberately vague
      because I am certain that e-mail based discussion is a deterrent to
      brainstorming. Second, I am convinced that we also need to address
      delivery issues, education being only part of the story. Obviously,
      this transcends the workshop and will require a Feyerabend project in
      the true sense of the word.
      --
      ----------------------------------------------------------------
      Theo D'Hondt
      Programming Technology Lab : Computer Science Department
      Faculty of Sciences : Brussels Free University
      Pleinlaan 2 / B-1050 Brussels / BELGIUM EUROPE
      mailto:tjdhondt@... http://prog.vub.ac.be/~tjdhondt
      Phone : +32-2-629 33 08 Fax : +32-2-629 35 25
      ----------------------------------------------------------------
    • Dave
      Education What happened to principle of computational/linguist variety? Theo, I very much share your concern about education. Some of the best educators in the
      Message 2 of 7 , Apr 16 5:55 PM
      • 0 Attachment
        Education What happened to principle of computational/linguist variety?

        Theo,

        I very much share your concern about education. Some of the best
        educators in the world have been forced to choose the same single
        language be it UML/Java/Ada/Basic ... to educate their students just
        because it is commercially popular. Their students have little or no
        chance to understand or respect the rich variety of computational and
        linguistic machinery of other visual or textual languages.

        Further some language/process/method cultures clearly discriminate
        against
        flexibility in favor of "important" computational values of type safety;
        performance; syntatic/semantic familiarity. Others stress flexibility;
        correctness at the expense of other features. No single language can
        capture the full spectrum of computation mechanism and experience.

        Social cultures form around languages - "real programmers don't script,
        they use a robust programming languages and leave the scripting to the
        blue collar programmers ..." "COBOL programmers are all old and
        dumb"...C programmers are hackers...

        Many programs don't even discuss the history of computation or
        programming languages. We as educators have a responsibility to at
        least educate our students cross techno-cultures --- "I saw an ER
        modeler talking to an Object Bigot last week ...."

        Many institutions are failing to do even this, let alone address the
        much more important socio-techical concerns raised by other posts.

        Dave

        > Many of us are educators and our educational concerns have an
        > important place in this discussion. I tend to share Dave Thomas'
        > preference for a focus on competence, maturity and knowledgeability
        > in development groups; this stands foursquare in opposition to the
        > process approach to software (which I fear is the mythical man-month
        > by another name). The latter is certainly inspired by the typical
        > profile of a computer professional today: his or her historical
        > perspective and historical consciousness is generally non existent.
        > We can only blame the educational system (and therefore ourselves)
        > for this state of affairs where education becomes training and is
        > driven by immediate concerns of the industry.
        >
      • Dave West
        I saw an ER modeler talking to an Object Bigot ... When this happens, the students love it!! I was doing a commercial seminar at a large software
        Message 3 of 7 , Apr 17 8:00 AM
        • 0 Attachment
          "I saw an ER modeler talking to an Object Bigot ..."

          When this happens, the students love it!! I was doing a commercial seminar
          at a large software development contract house (all attendees wwere were
          working software professionals) and one of my colleagues from the university
          was there. I was doing my usual 'object bigot' stuff (data is dead,
          databases are evil) and he could not stand it any longer and interjected.
          The rest of the afternoon was spent in a 'debate' with lots of questions and
          comments from the attendees.

          The attendees rated that seminar higher than any other ever offered at the
          company and we almost became a traveling roadshow - ala Timothy Leary and G.
          Gordon Liddy.



          Dave wrote:

          > Education What happened to principle of computational/linguist variety?
          >
          > Theo,
          >
          > I very much share your concern about education. Some of the best
          > educators in the world have been forced to choose the same single
          > language be it UML/Java/Ada/Basic ... to educate their students just
          > because it is commercially popular. Their students have little or no
          > chance to understand or respect the rich variety of computational and
          > linguistic machinery of other visual or textual languages.
          >
          > Further some language/process/method cultures clearly discriminate
          > against
          > flexibility in favor of "important" computational values of type safety;
          > performance; syntatic/semantic familiarity. Others stress flexibility;
          > correctness at the expense of other features. No single language can
          > capture the full spectrum of computation mechanism and experience.
          >
          > Social cultures form around languages - "real programmers don't script,
          > they use a robust programming languages and leave the scripting to the
          > blue collar programmers ..." "COBOL programmers are all old and
          > dumb"...C programmers are hackers...
          >
          > Many programs don't even discuss the history of computation or
          > programming languages. We as educators have a responsibility to at
          > least educate our students cross techno-cultures --- "I saw an ER
          > modeler talking to an Object Bigot last week ...."
          >
          > Many institutions are failing to do even this, let alone address the
          > much more important socio-techical concerns raised by other posts.
          >
          > Dave
          >
          > > Many of us are educators and our educational concerns have an
          > > important place in this discussion. I tend to share Dave Thomas'
          > > preference for a focus on competence, maturity and knowledgeability
          > > in development groups; this stands foursquare in opposition to the
          > > process approach to software (which I fear is the mythical man-month
          > > by another name). The latter is certainly inspired by the typical
          > > profile of a computer professional today: his or her historical
          > > perspective and historical consciousness is generally non existent.
          > > We can only blame the educational system (and therefore ourselves)
          > > for this state of affairs where education becomes training and is
          > > driven by immediate concerns of the industry.
          > >
          >
          >
          > To unsubscribe from this group, send an email to:
          > feyerabend-project-unsubscribe@yahoogroups.com
          >
          >
          >
          > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
        • Theo D'Hondt
          We re on exactly the same track. I became a kind of militant after last year s ECOOP educators workshop. On that occasion Kim Bruce told us that (as chairman
          Message 4 of 7 , Apr 19 12:33 AM
          • 0 Attachment
            We're on exactly the same track. I became a kind of militant after
            last year's ECOOP educators workshop. On that occasion Kim Bruce told
            us that (as chairman of some kind of high shot comittee on reform in
            IT education) he had had to accept that the first language in higher
            education should be an OO one, with explicit notions of using Java ...


            >Education What happened to principle of computational/linguist variety?
            >
            >Theo,
            >
            >I very much share your concern about education. Some of the best
            >educators in the world have been forced to choose the same single
            >language be it UML/Java/Ada/Basic ... to educate their students just
            >because it is commercially popular. Their students have little or no
            >chance to understand or respect the rich variety of computational and
            >linguistic machinery of other visual or textual languages.
            >
            >Further some language/process/method cultures clearly discriminate
            >against
            >flexibility in favor of "important" computational values of type safety;
            >performance; syntatic/semantic familiarity. Others stress flexibility;
            >correctness at the expense of other features. No single language can
            >capture the full spectrum of computation mechanism and experience.
            >
            >Social cultures form around languages - "real programmers don't script,
            >they use a robust programming languages and leave the scripting to the
            >blue collar programmers ..." "COBOL programmers are all old and
            >dumb"...C programmers are hackers...
            >
            >Many programs don't even discuss the history of computation or
            >programming languages. We as educators have a responsibility to at
            >least educate our students cross techno-cultures --- "I saw an ER
            >modeler talking to an Object Bigot last week ...."
            >
            >Many institutions are failing to do even this, let alone address the
            >much more important socio-techical concerns raised by other posts.
            >
            >Dave
            >
            >> Many of us are educators and our educational concerns have an
            >> important place in this discussion. I tend to share Dave Thomas'
            >> preference for a focus on competence, maturity and knowledgeability
            >> in development groups; this stands foursquare in opposition to the
            >> process approach to software (which I fear is the mythical man-month
            >> by another name). The latter is certainly inspired by the typical
            >> profile of a computer professional today: his or her historical
            >> perspective and historical consciousness is generally non existent.
            >> We can only blame the educational system (and therefore ourselves)
            >> for this state of affairs where education becomes training and is
            >> driven by immediate concerns of the industry.
            >>
            >
            >
            >To unsubscribe from this group, send an email to:
            >feyerabend-project-unsubscribe@yahoogroups.com
            >
            >
            >
            >Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/



            --
            ----------------------------------------------------------------
            Theo D'Hondt
            Programming Technology Lab : Computer Science Department
            Faculty of Sciences : Brussels Free University
            Pleinlaan 2 / B-1050 Brussels / BELGIUM EUROPE
            mailto:tjdhondt@... http://prog.vub.ac.be/~tjdhondt
            Phone : +32-2-629 33 08 Fax : +32-2-629 35 25
            ----------------------------------------------------------------
          • Uwe Zdun
            An interesting aspect is that in the technical reality the linguist diversity is creeping in again. In our experience, there is nearly no large scale system in
            Message 5 of 7 , Apr 19 1:45 AM
            • 0 Attachment
              An interesting aspect is that in the technical reality the linguist diversity
              is creeping in again. In our experience, there is nearly no large scale
              system in which we can't find multi-paradigms, multi-languages, etc. I've
              seen large "pure" Java (and C and C++) systems, built by people who believe
              scripting is evil, which contained little (awful) scripting languages. They
              would, of course, tell you that their systems are 100% statically typed OO.
              Since they haven't learned much about popular scripting languages before
              starting their design, their "little" scripting languages suffer from bad
              language design and many problems which have been solved long before.

              Another example are many of the XML based standards, such as RDF, XSL, etc.,
              which introduce linguist variety to the languages in which they are used as
              well.

              Yet another interesting aspect in this regard is that some languages foster
              diversity of dialects and extensions, notably Lisp, Tcl, XML ... from a
              certain point of view even C (many other languages are written in C and can
              be extended in C), others do not. I believe, only a few language elements are
              necessary for that: a way to treat program parts as data and vice versa, an
              intuitive extension mechanism to provide new first-class language
              entities, ...

              I think much of the linguist uniformity, we can observe in science, theories,
              education, product announcements, job announcements, etc., is rather
              artifical and does not reflect well the practices. However, in education it
              is, of course, harmful.

              --Uwe



              On Thursday 19 April 2001 09:33, you wrote:
              > We're on exactly the same track. I became a kind of militant after
              > last year's ECOOP educators workshop. On that occasion Kim Bruce told
              > us that (as chairman of some kind of high shot comittee on reform in
              > IT education) he had had to accept that the first language in higher
              > education should be an OO one, with explicit notions of using Java ...
              >
              > >Education What happened to principle of computational/linguist variety?
              > >
              > >Theo,
              > >
              > >I very much share your concern about education. Some of the best
              > >educators in the world have been forced to choose the same single
              > >language be it UML/Java/Ada/Basic ... to educate their students just
              > >because it is commercially popular. Their students have little or no
              > >chance to understand or respect the rich variety of computational and
              > >linguistic machinery of other visual or textual languages.
              > >
              > >Further some language/process/method cultures clearly discriminate
              > >against
              > >flexibility in favor of "important" computational values of type safety;
              > >performance; syntatic/semantic familiarity. Others stress flexibility;
              > >correctness at the expense of other features. No single language can
              > >capture the full spectrum of computation mechanism and experience.
              > >
              > >Social cultures form around languages - "real programmers don't script,
              > >they use a robust programming languages and leave the scripting to the
              > >blue collar programmers ..." "COBOL programmers are all old and
              > >dumb"...C programmers are hackers...
              > >
              > >Many programs don't even discuss the history of computation or
              > >programming languages. We as educators have a responsibility to at
              > >least educate our students cross techno-cultures --- "I saw an ER
              > >modeler talking to an Object Bigot last week ...."
              > >
              > >Many institutions are failing to do even this, let alone address the
              > >much more important socio-techical concerns raised by other posts.
              > >
              > >Dave
              > >
              > >> Many of us are educators and our educational concerns have an
              > >> important place in this discussion. I tend to share Dave Thomas'
              > >> preference for a focus on competence, maturity and knowledgeability
              > >> in development groups; this stands foursquare in opposition to the
              > >> process approach to software (which I fear is the mythical man-month
              > >> by another name). The latter is certainly inspired by the typical
              > >> profile of a computer professional today: his or her historical
              > >> perspective and historical consciousness is generally non existent.
              > >> We can only blame the educational system (and therefore ourselves)
              > >> for this state of affairs where education becomes training and is
              > >> driven by immediate concerns of the industry.
              > >
              > >To unsubscribe from this group, send an email to:
              > >feyerabend-project-unsubscribe@yahoogroups.com
              > >
              > >
              > >
              > >Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

              --
              Uwe Zdun
              Specification of Software Systems, University of Essen
              Phone: +49 201 81 00 332, Fax: +49 201 81 00 398
              zdun@..., uwe.zdun@...
            • Richard P. Gabriel
              ... I gave a talk at IBM a few weeks ago, and I included the following story. It s a bit of an exaggeration, but it is more true than not. In 1994, my company,
              Message 6 of 7 , Apr 20 3:29 PM
              • 0 Attachment
                At 20:55 -0400 4/16/01, Dave wrote:
                >Some of the best
                >educators in the world have been forced to choose the same single
                >language be it UML/Java/Ada/Basic ... to educate their students just
                >because it is commercially popular. Their students have little or no
                >chance to understand or respect the rich variety of computational and
                >linguistic machinery of other visual or textual languages.

                I gave a talk at IBM a few weeks ago, and I included the following
                story. It's a bit of an exaggeration, but it is more true than not.

                In 1994, my company, Lucid, folded and I started a hiatus that ended
                in 1998. I went from Lucid to ParcPlace where I decided to try
                managemnt in its own terms. After that I went back to school to get
                my mfa in poetry. I graduated in 1998 and then looked around at how I
                could do research again. My name had been made in Lisp - I founded
                Common Lisp and designed some of it, was a designer of CLOS, did a
                lot of parallel Lisp work - and aside from two hobbies - patterns and
                open source - I am known almost exclusively for Lisp stuff.

                I discovered that CS had decided that doing research in Lisp was
                forbidden: I could not find a place near where I live to do research
                in Lisp, I could not find any agency that would fund research in
                Lisp, and I found out that most of the publishing avenues for any
                work I did had dried up. Perhaps at MIT or Indiana or Rice I could do
                some work, but I don't (and can't) live there because of my family.

                This is very different from there simply being a trend away from a
                particular area. This felt as much like a deliberate act on the part
                of my field as anything could.

                So, I decided to stop caring - in my research - about helping the
                individual, but to concentrate instead on getting things improved in
                a global "mob" sense. I don't care whether any programmer suffers
                because what I care about is progress in the aggregate - like
                termites building arches or people building cathedrals or bizaars.

                Of course, this is exaggerated.

                -rpg-
              • fadrian@qwest.net
                ... part ... It may not be a conscious deliberate act, but the software Establishment does have a real message for all of it s stakeholders... To the
                Message 7 of 7 , Apr 20 4:34 PM
                • 0 Attachment
                  > This is very different from there simply being a trend away from a
                  > particular area. This felt as much like a deliberate act on the
                  part
                  > of my field as anything could.

                  It may not be a conscious deliberate act, but the
                  software "Establishment" does have a real message for all of it's
                  stakeholders...

                  To the programmer: I want you to continually upgrade your knowledge
                  to the "flavor of the day" so you can be a cheap, interchangeable cog
                  to help me change the "flavor of the day" so that...

                  To the consumer: You must upgrade to the "flavor of the day" because
                  it is NEW and it fixes defects in yesterday's "flavor of the day"! I
                  will charge you for the fixes because its not yesterday's "flavor of
                  the day" any more and it helps me bring out tomorrow's "flavor of
                  the day" having more defects.

                  To the educator: I want you to teach your pupils whatever
                  the "flavor of the day" will be when they graduate so they can help
                  me change the "flavor of the day"...

                  I call this "The Baskin-Robbins Model". Eat it today - we'll have
                  more for you to buy, tomorrow. In the end, the bottomless money
                  supply these feedback loops produce turn us all into whores and no
                  one gets to use last weeks "flavor of the day". Not that
                  tomorrow's "flavor of the day" will be any better - it's still all
                  crap at its core.

                  One way to break these cycles is simply to not stop the consumable
                  view of software. Right now we assume that software is to last no
                  longer than a year or so before defects or substrate changes force us
                  to upgrade. Software changes are linked to rapid hardware changes so
                  that whenever you buy new hardware, you need new substrate software
                  to support it, and if you have new substrate software... you're gonna
                  have to buy new application software to go along with it.

                  We know that it's possible to support legacy software better than we
                  do in the PC-based world. Look at IBM's zSeries and iSeries servers
                  (originally known as the S/390 and AS/400) - I can take a piece of
                  software from a machine three major versions back (that's about 15
                  years - they don't change the systems that quickly) and run it
                  unchanged on today's hardware. I'll probably be able to do that for
                  the next major release, too. Why? Because longevity is on eof the
                  things you buy with those machines.

                  How would a piece of software look if it were to be used by people
                  for the next 500 years? How about if it HAD to be maintained for the
                  next 500 years? Can we even imagine a language and/or substrate that
                  could be continually extended for 500 years?

                  I know that at least a precursor of the language exists - it's called
                  Lisp. It's maleable and can be "easily" processed to change when the
                  language changes. What about the substrate? Who cares. We've been
                  too close to it for too long, anyway.
                Your message has been successfully submitted and would be delivered to recipients shortly.